CONVERSATIONS
KEITH FERRELL
The Media Lab at M.I.T. just might be the hottest of all the world's hotspots for research into the future of television. In tact, future television is too narrow a phrase to describe all the areas Media Lab researchers are looking into.
"The focus of what we do and have been doing for 15 years," says Andrew Lippman, the Lab's associate director, "has to do with the merger of computing and image distribution. We variously explore that from the perspectives of education, technology, and entertainment."
Do those different perspectives change the nature of the research?
"Sometimes," Lippman says, "you think of that as interactive systems, and sometimes as high-definition television. But fundamentally and underneath those superficial distinctions, a lot of the work is very similar. It all addresses how you can smoothly merge what we've learned about computing and digital processing and image communications systems."
That smooth merger faces some obstacles, many of them imposed on television years ago. To understand the obstacles, Lippman compares TV's approach to images with the approach taken by computers.
"The focus on high-definition television, here," Lippman says, "can be encapsulated in one word: scalability. " For example, the Macintosh's small screen has 480 lines; the screen on a NeXT computer has 700 or 800 lines. An even bigger screen would carry even more lines. "The constant in these kinds of systems is the lines per inch, the density of lines," he explains. "And as you get a bigger screen, the density stays the same, and you get more lines."
That might sound logical, but it isn't the way television works.
"Only in television do you fix the number of lines," Lippman says. "And when you want a bigger display, [you] literally take those lines and spread them farther apart. That is counterintuitive; it just doesn't make any sense.
"We might do better in the next round if, instead of designing a TV system to be 500 lines, or 1000 lines, or some number like that, we optimized it for a system where [the number of] lines was not the number you specified, just like it isn't on a computer screen."
How would such a design change our household TVs?
"Your little TV under your kitchen counter might have 500 lines on it because it's only 4 inches high," Lippman says. "The one at the foot of your bed, which is a 19-inch set, might have 1000 lines. And the one that's on your wall—the lines on that will depend on your architect more than [on] the designer of your television."
Designer TV sets? Sure. "Likewise," Lippman says, "the shape will depend on your architect, as opposed to a Standards decision made in Washington. Kind of like theaters, where the shape of the screen is really the shape of the room. Maybe you should get your television set so that it's shaped to fit above the mantel, or along the wall. Scalability is the key."
The signal that scalable TVs would receive is as important as the sets themselves. "Having it be an end-to-end digital system at all phases is crucial because of the increasing number of digital channels and the increasing degrees of freedon that are afforded by digital representation," Lippman says.
Marrying scalable television sets with digital signals lays the ground-work for future television.
"If you can do that, then you have sown the seeds for future development of the medium," Lippman says. "You're no longer just sending out an analog 30-frame-per-second, or 50-frame, or 60-frame, motion picture; you're sending out data."
And that is the heart of the PC-TV connection. "Your [future] TV set has already got the processing needed simply to turn that data into a picture," Lippman explains. "That's 90 percent of a sophisticated computer. We can add the other 10 percent, give the TV some smarts, and let it start to help you make up the programs."
How dramatic a shift will this be for viewers accustomed to passive television viewing?
"You know, some [personalized TV technologies] are out there, in funny ways that you don't recognize," Lippman says. He cites a viewer's ability to program a VCR to tape programs for later viewing. "You're making up your own television, but it is a crude job," he says. "The control is pretty limited. All you can tell it is clock time and TV station to tape."
Tomorrow's television, thanks to computers, will offer more control. "If I can add content decisions to that, [for example:] ‘Watch the TV all the time for me, and if you see anything interesting about the world oil crisis, grab it, and I'll ask you later.’ That seems like a large step, but it's not," Lippman says.
Whatever the capabilities of upcoming television sets and systems, Lippman argues, the technologies must evolve on several fronts at once.
"The point is, can you make systems that, as they improve picture quality and give, strictly by fidelity, new degrees of freedom to existing uses like entertainment, can [those systems] also have the seeds of growth embedded within them?"